Essays on Household Behavior in the Recent Housing Cycle

Abstract

This dissertation studies a number of important issues brought up by the collapse of the housing market that began in the late 2000s. The first chapter takes up the issue of how policy can be crafted to assist ``underwater" households---those owing more on their house than what it is worth---an important population in a housing downturn. In particular, it provides an empirical evaluation of a mass refinancing initiative, the Home Affordable Refinance Program, that helped millions of households cut their monthly mortgage payments. This refinancing opportunity greatly reduced debt defaults for borrowers with poor credit scores, while simultaneously allowing them to spend more on durable consumption goods and pay down existing debts. The following two chapters take up the issue of sales volume; in particular, they look at why the number of homes sold rises dramatically during boom times and collapses so dramatically in busts. Chapter two of the dissertation empirically evaluates some explanations that have previously been proposed and finds that they are not nearly sufficient to explain the data. In particular, I find that mortgage frictions and nominal loss aversion among homeowners, while accurate descriptions of household behavior, are not particularly strong and, in an accounting sense, the movement in sales volume is driven by agents who do not face such issues. Chapter three proposes a novel mechanism. In particular, I argue that the booms experience an ``aggregate up-sizing" as many household move into nicer homes, generating high sales volume, but that, because of irreversibility of housing investment, there is no aggregate down-sizing in busts, so the market freezes and sales collapse. I then show that there is evidence both in aggregate housing market data and micro-level data that supports the importance of this mechanism.